It is our position that, for as long as is reasonably possible, planning be performed by measure of relative risk. We recommend this to manage expectations and, in particular, political ramifications of perceived certainty that is associated with the presentation of a numerical value.

The fundamental difference between absolute and relative measures is the use of a known baseline of risk. For example, one might assert that one in a thousand car trips results in fatality. This is an absolute measure, and carries with it certain expectations and management strategies, namely to reduce the number of car trips taken.

Its greatest weakness is the inherent exposure to challenge that comes about through its lack of qualifiers. There are obviously a number of direct and indirect causes of road use risk: Type of road, vehicle, speed, driver fatigue and so on. So to be correct, our initial statement needs to look like: One in a thousand car trips with a driver who hasn’t slept for 24 hours, on gravel roads in a car without ESC etc etc results in a fatality. It rapidly becomes less helpful, and one can be assured that all of the necessary qualifiers get lost in executive summaries across a bureaucracy.

Relative risk allows us to work from a baseline, when we may not even be sure about its precise location. Typically, medical alerts are given as relative risk, as there are simply too many drivers and underlying causes for medical situations to allow a meaningful absolute measure. A relative risk statement looks more like: It is twice as likely to be involved in a fatality, if the driver has not slept. It does not carry the implication that by ensuring all drivers have slept, there will be zero fatalities.

It is our experience that a number presented rapidly generates a life of its own. People can grasp a number, even though they may not understand what it purports to represent. It is exceedingly difficult to change a person’s perception once something as seemingly concrete as a numeric value is placed there.

When it comes to planning potential development sites, most true assessments of risk are relative. They are capable of correctly defining sites or configurations that present more risk than others. Their ability to qualify risk absolutely is questionable at best, even if this is due only to our evolving understanding of the actual causes of risk.

The implication is that we can reduce risk by modifying a proposal according to a single, identified potential of risk. Attempting to quantify the absolute reduction in risk that results from such a tactic opens an indefensible position that carries potentially significant implications. Ultimately, at some stage in the planning process we will be forced to take a categorical and absolute assessment of risk. Until such a time, it is our advice that all assessments be performed as relative measures and planning take the option that presents the least risk of two scenarios.